Road Edge Boundary Detection Using Multi-Sensor Prediction Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle mapping systems rely on manual detection of road edges, which is time-consuming, inaccurate, and resource-intensive, and fail to efficiently integrate new data or features, leading to incomplete and insufficient road edge boundary representations.
Innovation Solution
A computer-implemented method and system for automated road edge boundary detection using map data analysis, where prediction scores are generated based on sensor data and image parameters like LIDAR intensity, RGB values, and flatness images to accurately identify and connect road edge boundaries in a polyline format.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual detection methods are used for road edge identification, then detection accuracy can be maintained through human expertise, but the process becomes time-consuming and resource-intensive
Solution Approach 1:
The patent replaces manual mechanical detection with an automated computer vision system that uses machine learning models to detect road edges. The system processes sensor data and images automatically, eliminating the need for human operators to manually identify and mark road boundaries, thus reducing mapping time while maintaining detection accuracy through algorithmic precision.
Solution Approach 2:
The system enables the mapping process to be self-executing by using trained machine learning models that automatically analyze sensor data and generate road edge detections without human intervention. The automated pipeline processes images and sensor readings, identifies road boundaries, and updates maps autonomously, making the system self-sufficient for road edge detection tasks.
2Adaptability or versatility
If traditional mapping systems are used, then existing infrastructure can be leveraged, but the systems fail to efficiently integrate new data or features, leading to incomplete road edge boundary representations
Solution Approach 1:
The system implements a feedback mechanism where newly collected sensor data and images are continuously processed and compared against existing map data. The machine learning model evaluates prediction scores for potential road edge locations and updates the road edge boundary representations iteratively, ensuring that new data is efficiently integrated and road edge completeness is maintained and improved over time.
3Productivity
If automated detection systems are implemented, then mapping efficiency improves and manual time is reduced, but system complexity increases
Solution Approach 1:
The patent employs a universal machine learning framework that can process multiple types of sensor data and images through a single integrated system. The same detection pipeline handles various data formats and sensor inputs, reducing the need for separate specialized systems for different data types. This multi-functional approach improves mapping efficiency while managing system complexity through consolidation rather than proliferation of separate components.
Data Source
AI summary
Systems, devices, products, apparatuses, and/or methods for generating a road edge boundary for an edge of a road in an AV map for controlling an autonomous vehicle on a roadway by obtaining map data associated with a map of a geographic location including a roadway associated with one or more locations of one or more vehicles in the roadway during one or more traversals of the roadway, determining one or more prediction scores based on the map data, including one or more predictions of whether the plurality of elements include road edge boundary locations, and generating in the map a road edge boundary based on the one or more prediction scores.